Post

🧾 Dev Log – From Theory to Data Analytics

🧾 Dev Log – From Theory to Data Analytics

⏳ The Evolution: Late Summer to Now

Since August, I’ve measured my progress not in months, but in friction:

  • Fighting with Jupyter vs. Terminal logic.
  • Reviving an old laptop with Debian just to have a dedicated dev environment.
  • Watching GitHub Actions fail until they finally didn’t.

The goal was simple: Stop living in static Excel formulas and start building reproducible data systems. What followed was a messy, rewarding transition from “theory” to a portfolio grounded in reality.


🧱 The Build: Portfolio Highlights

1️⃣ Finance Foundations

Before pivoting to pure analytics, I built tools to automate the technical debt of finance:

  • Custody NAV Calculator: Python-based automation for daily net asset values.
  • Portfolio Risk Report: My first engine for wrapping complex logic into structured outputs.
  • Thesis Backtesting: Testing ETF vs. SPX strategies using historical data.

2️⃣ iGaming & Behavioral Analytics

This was the shift from “rows of data” to “customer journeys.”

  • Retention & Churn: Built SQL logic to define cohorts and LTV.
  • Product Thinking: Moving beyond SELECT * to answer questions about player engagement and risk.

3️⃣ Banking Customer Intelligence

I scaled a raw banking dataset from 5K to 100K records using Python, then architected a PostgreSQL schema to analyze it.

  • The Goal: Identify high-value customers vs. churn risks.
  • The Result: Professional Tableau dashboards built on structured schemas, not just “pivots.”

4️⃣ Louisville Metro Public Payroll

My most ambitious project to date, using 40,000+ real-world salary records.

  • The Tech: PostgreSQL (Indexes, CTEs, Window Functions) + Tableau Public.
  • The Insight: Mapped overtime intensity and pay inequality across five years of city data.
  • The Shift: I stopped “inventing” companies and started solving problems with messy, public data.

🧭 What’s Next for 2026?

Technical Depth

  • Statistical Modeling: Implementing hypothesis testing, and logistic regression.
  • Python: Moving from Notebooks to clean, scheduled ETL scripts.

Freelance Projects

I am shifting toward projects from clients:

  • Business Dashboards: Turning messy invoices into clear output.
  • Data Audits: Templates for migrating business logic from “spreadsheet chaos” to structured databases.

⚡ Connect

GitHub Profile Email LinkedIn

© 2025 Pietro Di Leo. One commit at a time.

This post is licensed under CC BY 4.0 by the author.